Fig 1. A typical thermoelectric (TE) energy harvesting system consists of five key components—the TE generator, the heatsink, voltage regulation, charge management and energy storage, and power/load management—as well as a heat source.
Fig 2. The thermal circuit can be seen as an equivalent electrical circuit where the heat flow is represented by current and the temperature is represented by voltage. Each component in the thermal circuit has an associated thermal resistance.
Fig 3. The PN couple is the basic building block of a TEG device. The P and N elements are connected electrically in series and thermally in parallel to produce current flow.
Fig 4. TEGs can be used to power wireless sensor systems in a variety of applications, such as the detection, diagnosis, and prognosis of problems in turbine engine bearing assemblies.
With the adoption of new wireless protocols and improvements in transceiver technology, wireless networks for distributed sensors are quickly becoming commonplace. As the power usage increases, energy scavenging can effectively provide a virtually perpetual energy source that does not require frequent maintenance. Thermoelectric (TE) energy harvesting is an excellent choice when modest heat sources are available and when light or vibration energy is unavailable.
A TE energy harvesting system consists of five major components in addition to a heat source (Fig. 1). The first two items in this list, the TE generator (TEG) and the heatsink, can be viewed as the thermal circuit. Items three, four, and five—the voltage regulation, the charge management and energy storage, and the power/load management—are part of the electrical circuit in conjunction with the TEG.
TE energy conversion requires special attention to both circuits to provide optimal performance. Put another way, it is often necessary to design and build the system for a known set of conditions to provide the most efficient energy conversion.
The Thermal Circuit
The thermal system is viewable as an equivalent electrical circuit wherein the heat flow is represented by current and temperature by voltage. Each component in the thermal circuit has an associated thermal resistance (Fig. 2).
In addition, interfaces between components have non-negligible thermal resistances that must be accounted for. Similar to a battery connected to an electrical load, the maximum power converted from heat into electrical power in the thermoelectric generator occurs when the thermal resistance of the heat source and heat exchangers matches that of the TEG.
Heat Sources
Heat sources can take on many different forms including hot surfaces such as pipes, furnaces, engine blocks, exhaust gases, direct sunlight, and even the human body. Sources can also vary considerably in maximum temperatures available and variability over time.
Additionally, heat sources have associated internal thermal resistance and maximum heat flux capabilities. The source temperature variability and thermal resistance have to be taken into account in the design of the energy harvester.
The TE Generator
The direct conversion of heat energy into electrical energy can be accomplished through the Seebeck effect, in which an induced heat flow through an appropriately designed thermoelectric device produces a voltage and current. The PN couple, which is the basic building block of a TEG device, comprises a single pellet of P-type and N-type thermoelectric material each that are connected electrically in series (Fig. 3).
Heat carries most carriers from one junction to the other, producing a current and voltage. By placing many PN couples in series electrically and in parallel thermally, a typical TE module can be constructed that generates a voltage proportional to the temperature differential across the elements.
Heatsinks
The power generated by a TEG is proportional to the square of the delta T (ΔT) across the TEG module. To enable the constant flow of heat required to sustain the ΔT, a heatsink or exchanger is required. Heatsinking, or rejection, can be accomplished in numerous ways, i.e., through direct conduction to a large thermal mass, by liquid cooling, spray cooling, or phase change, or simply through convection to the air using a traditional heatsink.
Many thermoelectric energy harvesting applications rely on air cooling and therefore conventional heatsinks. In addition, these sinks are typically the lowest-cost solution and easiest to implement. The electrical system includes the TEG, the voltage regulator, charge management circuitry, energy storage, and power (load) management circuitry (Fig. 1, again).
Voltage Regulation
Most electronics require more than 1 V to operate, and many require greater than 3 V. Since the voltage output of a TEG is proportional to the temperature differential ΔT across the TEG, large values of ΔT are desired for most applications. In many applications, the ΔT available is fairly small, meaning the output voltage of the device may not exceed the minimum requirement.
Variation in the temperature of the heat source can also lead to unstable voltage output. To address these issues, it is necessary to boost (and regulate) the voltage to the minimum, usable, constant value. The challenges associated with this regulation lie in getting enough starting voltage to bootstrap the convertor and getting enough output power from the convertor given the conversion efficiency. Boost converters (greater than 400-mV input) and micropower converters (greater than 30-mV input) can address this need.
Charge Management And Energy Storage
There are two basic approaches to storing energy in an energy harvesting unit: electrochemical cells (batteries) and capacitors. Batteries are capable of high-energy storage densities compared to capacitors. A lithium-ion (Li-ion) battery can store up to 100 to 250 Wh/kg of energy, and nickel-metal-hydride (NiMH) batteries are typically on the order of 30 to 100 Wh/kg. This compares to 3 to 5 Wh/kg for supercapacitors. Batteries also exhibit less leakage current than capacitors.
Capacitors, particularly supercapacitors, have high power densities compared to batteries. This means they can provide high bursts of current for short periods of time. Power densities for Li-ion batteries range from 250 to 1000 W/kg compared to supercapacitors, which can be as high as 8000 W/kg. In cases where high-energy storage is required, but short bursts of energy are also needed, a capacitor can be placed in parallel with the cell.
Thin-film batteries offer a new generation of technology that provides high-energy storage (capacity, C, ranging from 0.050 to 2.5 mAh) in thin, surface-mountable configurations. This makes integration of these cells ideal on small circuit boards.
Thin-film battery manufacturers have their own charge management circuits built into some of their products. These charge controllers are capable of lower standby power consumption in the 10-µW range, and they’re better suited for micropower energy harvesting systems.
Power/Load Management
The load management circuit is used to automatically switch the load between the regulated voltage supplied by the TEG and the battery depending on the thermal conditions and status of the regulator output, i.e., the TEG and the discharge level of the battery. Battery protection circuits often are also incorporated to ensure that the battery is not overcharged or over-discharged.
The load management circuit also can be used to match the load seen by the thermoelectric generator to maintain the energy harvesting system closer to the maximum power transfer point. However, this requires a specially designed boost input stage and the ability to do load sharing if necessary.
Example Application
To illustrate the application of thermal energy harvesting in a distributed wireless sensor network, Figure 4 depicts the integration of a TEG with a wireless sensor system designed to monitor the health of bearings in turbine engines. The system allows for the detection, diagnosis, and prognosis of problems occurring within an engine bearing assembly.
Some of the most mechanically stressed components within turbine engines include the bearing assemblies. Real-time monitoring of temperature, vibration, strain, and pressure can provide critical information on the health of bearings for both aircraft turbines and power generation systems.
By analyzing data from these systems, engine maintenance will not be required until the sensor shows it is necessary, rather than at regularly scheduled intervals, saving a great deal of time and money. Real-time detection and diagnosis of faults in aircraft turbine bearings not only can alert the pilot that there is a problem in the engine, it also can inform the maintenance crew where the actual fault is occurring.
Both power generation and aircraft turbine designers can also benefit from data collected during continuous field operation of their engines, allowing for design improvements in the safety, reliability, and efficiency of future systems.
The TEG can generate about 20 mW to 30 mW from the heat of normally operating turbine engine bearings. To maintain the proper temperature differential, the system takes advantage of oil spray cooling in the jet engine. The sensor communicates wirelessly to receivers that are distributed within the aircraft that aircraft crews or pilots use to monitor the conditions of each bearing.
The solution eliminates the cost of battery replacement for the sensor system. It also aids in the integration and retrofitting of existing turbine engines.